NNTP and thunderbird-related lists

Dan Mosedale dmose at mozilla.org
Fri Jul 30 18:50:47 UTC 2010


  On 7/29/10 5:49 PM, Ben Bucksch wrote:
>
>
> Below, I am assuming that web archives do not exist, due to their bad 
> and cumbersome UI. I don't want to use them.
>
> [...]
>
> If I am not subscribed, I don't have access to the posts, plain and 
> simply. I cannot post myself, either. And that's because the protocol 
> is push, and has no pull. Anything before the moment I subscribe is 
> not available to me, and anything that didn't happen yet is obviously 
> not available either, meaning that at the moment I first look at the 
> mailing lists, I have zero posts to look at, and it's because of the 
> push nature.
>
> With newsgroups, I can fetch a certain history, any time I like, 
> because they are pull.
>
Ah, ok; this explanation helped.  I get what you mean now.
>> My non-understanding of how pull vs. push is relevant to the above 
>> paragraph is because I subscribe to a bunch of mailing lists which I 
>> filter into folders, and some of them I read in very much the style 
>> you describe: I go look at them once a week or once a year.  Would 
>> that not work for you?
>
> A, you assume that I had discovered the mailing list a while ago, and 
> subscribed back then. That ignores the first time I encounter the 
> group, though.
Quite right, it doesn't solve that problem.
> B, at the time of first encounter, I am probably not sure whether I 
> want the list/group. And I have no way to read the history 
> conveniently (see above). Unless I have good reason, I will probably 
> not subscribe. Once I have good reason, I cannot read the posts that 
> pose that good reason and cannot reply, because there's no history.
Cannot is a bit of a strong word here, given that the archives do 
exist.  I get, however, that it's a worse user-experience for 
interacting with history posting, and that this will probably deter some 
subscribers.
> C, I find it totally pointless and resource-wasting to subscribe to a 
> mailing list that I intend to look at only once a month or year. If I 
> look at it only once a year, I probably don't know whether I'll *ever* 
> look at it again. Why should I continuously receive messages when I 
> don't know whether I'll ever read them? It's most obvious with the 
> linux-kernel mailing list (LKML). I am occasionally interested, I read 
> maybe a few posts every few months, when somebody points out something 
> funny or of particular interest. I would not subscribe to it, however, 
> because it's a high-volume list and I'd get hundreds of msgs per day. 
> That's an extreme example, but shows the tradeoff. Most tradeoffs are 
> closer, but once the value ratio goes below e.g. 50% or 30% or 10%, 
> many people feel they should not receive the whole stream. Even if 
> they are definitely interested in certain, occasional threads.
Agreed; using an email list for these sorts of discussions is not an 
optimal use of resources, and that might deter some people from subscribing.
> In other words, mailing lists are not just "High barrier/effort to 
> entry", they are simply "*No* entry unless you're a regular". Not 
> everybody wants to be a regular / club member. I want to just visit 
> the restaurant I feel like on that evening, without being forced to 
> subscribe as member and then wait a few days. I want to visit many 
> different restaurants without being a "member" everywhere.
> (OK, the comparison lacks, because to join on tb-planning discussion, 
> you better be a regular. But not to just read. For tb-enterprise, you 
> shouldn't need to be a regular, but you should definitely skim over 
> the history before posting!)
So it sounds like you're saying that, if one ignores the existence of 
the archives, it's more than just a high barrier to entry.  Got it.

Dan



More information about the tb-planning mailing list